1) Define the mechanisms by which the specificity and fidelity of miRNA biogenesis are controlled. Drosha, a conserved RNase III enzyme, plays a critical role in maturation of microRNAs (miRNA). Drosha interacts with a dsRNA-binding protein encoded by DiGeorge syndrome critical region 8 (DGCR8) to form the Microprocessor complex, which releases hairpin-shaped miRNA precursors (pre-miRNA) from larger RNA polymerase II (Pol II)-driven precursor transcripts in the nucleus. Regulation of miRNA function happens at the first step of its biogenesis. A number of protein factors, including Lin28, p53, Smads and SRSF3 (SRp20), which interestingly also function as tumor suppressors or oncogenes, are suggested to act as modulators of activity and specificity of Drosha-mediated pri-miRNA cleavage. Reduced enzymatic activity of Drosha has been described in various malignancies. In the past year, we found that Drosha was under diverse regulations in the post-transcriptional level. In both Hela and HEK293 cells, we identified then successfully cloned two novel in-frame isoforms generated by alternative splicing. Using junction primers and Drosha specific Tagman probe, we measured the mRNA level of these isoforms by quantitative RT-RCP. To our surprise, more than 50% of Drosha mRNAs were under posttranscriptional regulation, generation isoform proteins in comparable amount to that of full-length Drosha. In addition, we created Drosha KO cell lines and a set of luciferase based reporters, which render us the opportunity to detect potential novel functions of these isoforms in vivo. Next generation sequencing analysis and functional assay are current ongoing. 2) Create a novel model demonstrating how the miRNA decay is regulated by end modifications. Imprecise cleavages during biogenesis and post-maturation modifications generate miRNA variants termed isomiRs, whose sequences are slightly different from canonical miRNA at either 5' and/or 3' ends. While many efforts have focused on studying the alteration in isomer expression levels during cancer development, little is known about how miRNA 3' modifications are regulated. Their impacts on miRNA function remain largely elusive. To further investigate miRNA modification, we over-expressed hsa-miR-27a in HEK293 cells. RNA samples were collected at five time points (12hr, 24hr, 36hr, 48hr, 60hr) post-transfection. Level of isomiRs were measured by deep sequencing and validated by Northern blotting. To our surprise, more than half of the miR-27a is subject to 3' end modifications in all samples measured. Majority of the modifications are losing the last C (trimming), gaining an additional U (tailing) or combination of both. While the isomiR profile of endogenous hsa-miR-27b remained unchanged, pattern of 3' modification on over-expressed hsa-miR-27a altered over time: Trimming increased whilst the percentage of tailing raised at the first 24hr and then gradually reduced. Those observations were consistent with two additional miRNAs (hsa-miR-23a and hsa-miR-134), suggesting that the 3' modifications are not random but rather tightly regulated. Notably, when level of overexpressed miR-27a peaked around 60hr post-transfection, its isomer profile became similar as endogenous miR-27a, despite the fact that the level of former was over a hundred fold more than the latter. This suggests the 3' end modification is not a consequence of miRNA accumulation, but rather plays an active role on maintaining the miRNA hemostasis by stabilizing miRNA or promoting decay. This study furthered our understanding on isomiR dysregulation in cancer, providing new insights on the regulation and function of isomiRs. It also shed lights on designing new approaches to manipulate miRNA level/function in cancer treatments. 3) Develop new strategies for designing potent DNA-directed RNAi for cancer treatment. DNA-directed RNAi (shRNA expressed from plasmids) is more desirable than traditional synthetic siRNAs in many applications. It is preferred or required in genetic screens and specific RNAi approaches in gene therapy settings. However, the application is hampered due to unwanted off-target effects, a major source of which originates from heterogeneous products of shRNA processing in vivo. Our work on Dicer processing has established the loop-counting rule, which laid the groundwork in designing Pol III driven pre-miRNA-like (first generation) shRNAs free of heterogeneous processing. Together with knowledge of Drosha processing gained, we are developing strategies in designing safer pri-miRNA-like (second generation) shRNAs, which can be expressed with a Pol II promoter and are amenable to more transcription controls.